Forecasting Is Fallible, yet Necessary

There are two types of forecasting.

1) Trying to estimate the effect of policy and risks.

2) Trying to forecast what actually happens in the future for real.

Most of macroeconomic forecasting is forecasting of the first kind. Central banks forecasters use modeling to help decide the policy that nudges the economy right direction. They want to forecast what is the effect of their actions in the near future, not the actual future influenced by events that are outside their control.

Financial forecasting is similar. Predicting how the future unfolds would be nice, but predicting the expected value is more realistic. Just like poker players can't see what happens next but they can still get it right, risk analysis based on modeling future scenarios looks like forecasting but it's not exactly forecasting in the sense of predicting what actually happens.